| *** aselius has quit IRC | 00:29 | |
| *** yamahata has quit IRC | 01:48 | |
| *** diablo_rojo has quit IRC | 01:50 | |
| *** edmondsw has joined #openstack-meeting-cp | 01:51 | |
| *** iyamahat has quit IRC | 01:54 | |
| *** edmondsw has quit IRC | 01:55 | |
| *** markvoelker_ has quit IRC | 02:41 | |
| *** zhipeng has joined #openstack-meeting-cp | 02:54 | |
| *** zhipeng has quit IRC | 03:06 | |
| *** edmondsw has joined #openstack-meeting-cp | 03:39 | |
| *** edmondsw has quit IRC | 03:44 | |
| *** markvoelker has joined #openstack-meeting-cp | 04:42 | |
| *** iyamahat has joined #openstack-meeting-cp | 05:10 | |
| *** iyamahat has quit IRC | 05:11 | |
| *** iyamahat has joined #openstack-meeting-cp | 05:11 | |
| *** gouthamr has quit IRC | 05:13 | |
| *** markvoelker has quit IRC | 05:17 | |
| *** rarcea has joined #openstack-meeting-cp | 05:18 | |
| *** edmondsw has joined #openstack-meeting-cp | 05:28 | |
| *** edmondsw has quit IRC | 05:32 | |
| *** hemna_ has quit IRC | 06:01 | |
| *** markvoelker has joined #openstack-meeting-cp | 06:14 | |
| *** markvoelker has quit IRC | 06:47 | |
| *** MarkBaker has joined #openstack-meeting-cp | 07:12 | |
| *** coolsvap has joined #openstack-meeting-cp | 07:13 | |
| *** edmondsw has joined #openstack-meeting-cp | 07:16 | |
| *** edmondsw has quit IRC | 07:20 | |
| *** markvoelker has joined #openstack-meeting-cp | 07:44 | |
| *** MarkBaker has quit IRC | 07:45 | |
| *** markvoelker has quit IRC | 08:16 | |
| *** iyamahat has quit IRC | 08:29 | |
| *** MarkBaker has joined #openstack-meeting-cp | 08:36 | |
| *** MarkBaker has quit IRC | 08:56 | |
| *** MarkBaker has joined #openstack-meeting-cp | 09:03 | |
| *** edmondsw has joined #openstack-meeting-cp | 09:04 | |
| *** edmondsw has quit IRC | 09:09 | |
| *** markvoelker has joined #openstack-meeting-cp | 09:13 | |
| *** markvoelker has quit IRC | 09:46 | |
| *** edmondsw has joined #openstack-meeting-cp | 10:52 | |
| *** edmondsw has quit IRC | 10:57 | |
| *** markvoelker has joined #openstack-meeting-cp | 11:44 | |
| *** edmondsw has joined #openstack-meeting-cp | 11:50 | |
| *** MarkBaker has quit IRC | 12:13 | |
| *** MarkBaker has joined #openstack-meeting-cp | 12:14 | |
| *** markvoelker has quit IRC | 12:16 | |
| *** markvoelker has joined #openstack-meeting-cp | 12:32 | |
| *** rarcea has quit IRC | 12:49 | |
| *** rarcea has joined #openstack-meeting-cp | 13:01 | |
| *** gouthamr has joined #openstack-meeting-cp | 13:27 | |
| *** edmondsw has quit IRC | 13:40 | |
| *** edmondsw has joined #openstack-meeting-cp | 13:56 | |
| *** hemna_ has joined #openstack-meeting-cp | 15:21 | |
| *** diablo_rojo has joined #openstack-meeting-cp | 15:21 | |
| *** diablo_rojo has quit IRC | 15:44 | |
| *** aselius has joined #openstack-meeting-cp | 15:49 | |
| *** mriedem has joined #openstack-meeting-cp | 15:55 | |
| ildikov | #startmeeting cinder-nova-api-changes | 16:00 |
|---|---|---|
| openstack | Meeting started Thu Aug 24 16:00:19 2017 UTC and is due to finish in 60 minutes. The chair is ildikov. Information about MeetBot at http://wiki.debian.org/MeetBot. | 16:00 |
| openstack | Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. | 16:00 |
| *** openstack changes topic to " (Meeting topic: cinder-nova-api-changes)" | 16:00 | |
| openstack | The meeting name has been set to 'cinder_nova_api_changes' | 16:00 |
| ildikov | johnthetubaguy jaypipes e0ne jgriffith hemna mriedem patrickeast smcginnis diablo_rojo xyang1 raj_singh lyarwood jungleboyj stvnoyes | 16:00 |
| hemna | hi | 16:00 |
| jungleboyj | @! | 16:00 |
| mriedem | o/ | 16:00 |
| _pewp_ | jungleboyj ヽ(´･ω･`)､ | 16:00 |
| smcginnis | o/ | 16:00 |
| jungleboyj | Though also in another meeting too. :-) | 16:00 |
| ildikov | jungleboyj: thanks for keeping an eye on this one too :) | 16:01 |
| jungleboyj | ildikov: I do what I can. :-) | 16:01 |
| ildikov | jungleboyj: :) | 16:01 |
| stvnoyes | o/ | 16:01 |
| ildikov | ok, let's start | 16:02 |
| ildikov | current open reviews: https://review.openstack.org/#/q/topic:bp/cinder-new-attach-apis | 16:02 |
| smcginnis | https://review.openstack.org/#/q/topic:bp/cinder-new-attach-apis+status:open | 16:02 |
| smcginnis | ;) | 16:02 |
| ildikov | smcginnis: my bad :) | 16:03 |
| smcginnis | Just a more concise view. :) | 16:03 |
| ildikov | mriedem: the two on the top of the list are small changes to add an attachment_complete call and to tweak yet once more on the attachment_ref, mainly the connection_info part | 16:03 |
| ildikov | mriedem: those can be reviewed already | 16:04 |
| ildikov | for the attach patch we still need a client release to get a clean test run on the gate | 16:04 |
| ildikov | locally things seem to work | 16:04 |
| ildikov | except one thing that stvnoyes has just found with 'cinder show volume' | 16:05 |
| smcginnis | ildikov: Latest on a new client release - looks like that will need to be next week. | 16:05 |
| ildikov | with this patch we seem to use volume_id as opposed to attachment_id: https://github.com/openstack/python-cinderclient/commit/63ac82a55489d55246da939b5ae60b8a65fb9ec7 | 16:05 |
| ildikov | smcginnis: is that still issues with stable? | 16:06 |
| ildikov | smcginnis: we might need to fix one more thing in the client, so from that perspective it sounds ok | 16:06 |
| ildikov | smcginnis: I just wonder what else we have on the table | 16:07 |
| smcginnis | ildikov: Yeah, the lib freeze will need to go until everything is complete for pike. | 16:07 |
| ildikov | smcginnis: I guess we should be confident with end of next week then? | 16:07 |
| smcginnis | ildikov: Better be, otherwise something went really wrong. ;) | 16:08 |
| mriedem | i'd like more details in the commit message for https://review.openstack.org/#/c/493324/ | 16:08 |
| ildikov | smcginnis: my thinking as well, just wanted to be sure :) | 16:08 |
| mriedem | ildikov: also, can you re-propose the spec for queens and make any necessary updates? https://review.openstack.org/#/c/373203/ | 16:09 |
| *** kbyrne has quit IRC | 16:09 | |
| ildikov | mriedem: sure, will do | 16:09 |
| mriedem | i guess i never got https://review.openstack.org/#/c/459134/ in | 16:09 |
| ildikov | mriedem: yeah, I planned to, wanted to double check with you | 16:09 |
| ildikov | mriedem: I will add that and I should add attachment_complete to it too | 16:10 |
| *** kbyrne has joined #openstack-meeting-cp | 16:10 | |
| mriedem | i've lost context on what my amendment is even about anymore | 16:10 |
| mriedem | it was basically trying to capture discussion for something we'd need to consider when supporting multi-attach i think | 16:11 |
| ildikov | it was about keeping begin_detaching with the new flow too, so we have the call in the API to put the volume into 'detaching' state as I remember | 16:12 |
| ildikov | which is pretty much needed for multi-attach as we don't want two detaches going at the same time | 16:12 |
| *** coolsvap has quit IRC | 16:13 | |
| jgriffith | ildikov why not? | 16:13 |
| jgriffith | We've again kinda lost sight of the whole point of discreet and independent attachment objects I think | 16:14 |
| jgriffith | heck, at this point I'm not even sure I understand it any more :) | 16:14 |
| ildikov | jgriffith: well, if we don't change the volume state that can be attach and etach as well that goes in parallel and I'm not sure that's a good idea | 16:15 |
| ildikov | jgriffith: and the volume state machine doesn't support it either | 16:15 |
| jgriffith | probably not a discussion for right now, but that whole thing needs to go away/stop | 16:15 |
| ildikov | so for now it's better not to create a mess :) | 16:15 |
| jgriffith | the attachment needs to be the source of truth | 16:15 |
| ildikov | it sounds good for the long run | 16:16 |
| ildikov | but we can't start those changes in Nova | 16:16 |
| jgriffith | understood | 16:16 |
| jgriffith | but don't try and cheat and force in a half assed multi-attach solution either :) | 16:16 |
| mriedem | ildikov: how about we just re-propose the spec and leave out https://review.openstack.org/#/c/459134/ | 16:17 |
| mriedem | we can roll that into the queens spec later if we want | 16:17 |
| jgriffith | anyway, not sure it's really relevant quite yet... sorry | 16:17 |
| jgriffith | don't want to derail | 16:17 |
| mriedem | right let's avoid the noise | 16:17 |
| mriedem | and just ignore https://review.openstack.org/#/c/459134/1 | 16:17 |
| ildikov | jgriffith: we will talk about the multi-attach bits on the PTG, I hope :) | 16:17 |
| ildikov | mriedem: works for me | 16:17 |
| ildikov | mriedem: thanks for the hint | 16:18 |
| ildikov | jgriffith: BTW, did you see my above comment on cinder show volume not returning the correct attachment_id? | 16:18 |
| jgriffith | No | 16:18 |
| ildikov | jgriffith: stvnoyes has just found out as he uses that for one of his tempest tests | 16:18 |
| smcginnis | mriedem: Do we have a good time for a joint PTG session? | 16:18 |
| smcginnis | jungleboyj: ^ | 16:18 |
| mriedem | for a good time... | 16:18 |
| mriedem | like on a bathroom wall? | 16:19 |
| mriedem | smcginnis: anytime wed-fri is fine | 16:19 |
| mriedem | we just have an etherpad with proposed stuff right now, nothing in order | 16:19 |
| ildikov | jgriffith: so this is apparently the volume_id for some reason: https://github.com/openstack/python-cinderclient/commit/63ac82a55489d55246da939b5ae60b8a65fb9ec7#diff-799a44cd44f14a12e4f1d2a8b56ab63aR39 | 16:19 |
| smcginnis | mriedem: Same for us. | 16:19 |
| jungleboyj | :-) mriedem We should coordinate that. | 16:19 |
| mriedem | i hope we're not in the same wedding ballrooms as last time | 16:20 |
| ildikov | smcginnis: thanks for bringing it up, so I don't forget :) | 16:20 |
| smcginnis | ildikov: I've been meaning to for a while, so I figured I better while we're all here. :) | 16:20 |
| jungleboyj | mriedem: Do you have a preference. Maybe Thursday some time? | 16:20 |
| mriedem | jungleboyj: that works | 16:20 |
| jgriffith | wedding? who's getting hitched? Free food and drink.. YAY | 16:20 |
| mriedem | thursday morning | 16:20 |
| smcginnis | ++ | 16:20 |
| jungleboyj | jgriffith: ++ | 16:20 |
| smcginnis | Was just going to suggest morning so there's follow on time. | 16:20 |
| mriedem | i just want a room with a ceiling that's <20 feet | 16:20 |
| jungleboyj | mriedem: That sounds good. | 16:20 |
| ildikov | sounds good | 16:21 |
| jungleboyj | jgriffith: ildikov You guys will get everyone up to speed on where the API rework and impacts during that time? | 16:21 |
| jungleboyj | mriedem: Is there interest on the Nova side there? I know that the Cinder team needs to be updated. | 16:21 |
| * smcginnis has to drop off for a bit for solo parent duties | 16:22 | |
| ildikov | jungleboyj: that's an option, but it would be more high level if we have both teams there | 16:22 |
| mriedem | i'd be happy if more nova people were involved | 16:22 |
| ildikov | mriedem: +2 +A | 16:23 |
| mriedem | if johnthetubaguy is going to be at the ptg he'd need to be updated | 16:23 |
| mriedem | otherwise maybe gibi will get the mantle | 16:23 |
| jungleboyj | johnthetubaguy: You going to be there | 16:23 |
| * jungleboyj whispers "please say yes, please say yes" | 16:23 | |
| ildikov | I think he has a ton of concerns without update too :) | 16:23 |
| mriedem | yeah anyway | 16:23 |
| diablo_rojo_phon | He is going to be there. | 16:23 |
| jungleboyj | Ok, that is good news. | 16:24 |
| mriedem | what next | 16:24 |
| * mriedem still has rc2 bugs to fix | 16:24 | |
| ildikov | mriedem: yeah, I gave gibi an idea on how annoying I can be with review requests in the second line after 'congrats'... :) | 16:24 |
| ildikov | mriedem: not much more on the Nova side until we don't have the cinderclient fixed and released | 16:24 |
| mriedem | ok | 16:25 |
| ildikov | jgriffith: so as for volume show, it seems to be an easy client side fix | 16:25 |
| mriedem | maybe i'll throw https://review.openstack.org/#/c/463987/ in next week | 16:25 |
| mriedem | stvnoyes: is ^ still happy with the new flow? | 16:25 |
| mriedem | or i guess we can't test that in upstream CI until cinderclient fixes are released | 16:25 |
| stvnoyes | working thru that cinder client issue atm | 16:26 |
| mriedem | ok, because the stack of api changes at https://review.openstack.org/#/c/330285/ depends on the cinderclient fixes | 16:26 |
| mriedem | so i guess i'll hold off | 16:26 |
| ildikov | jgriffith: it's weird why we have the volume_id as the value of 'id' under 'attachments' in the volume though, at least for my taste | 16:27 |
| ildikov | jgriffith: maybe hemna has the history of that | 16:27 |
| ildikov | mriedem: well, things mostly look good locally, but I guess it sounds less convincing anyway | 16:27 |
| jgriffith | ildikov there's a lot of garbage in there frankly | 16:28 |
| jgriffith | like storing attach-status in 3 different places etc; I'll look at the volume-show stuff you mentioned and see what's up | 16:28 |
| mriedem | the attachment_id in the compute API for attached volumes has always been the volume_id i think | 16:29 |
| stvnoyes | jgriffith- in cinderclient.v2.shell.py._translate_attachments, the attachment_id is replaced with attachment['id'] which happens to be the volume id | 16:29 |
| jgriffith | mriedem I believe you're correct on that | 16:29 |
| mriedem | https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/volumes.py#L226 | 16:29 |
| ildikov | jgriffith: I think for now we can just use the attachment_id in the translate function in the client | 16:29 |
| mriedem | super legacy | 16:29 |
| stvnoyes | the correct attachment id is in 'attachment_id' | 16:29 |
| stvnoyes | not id | 16:30 |
| jgriffith | stvnoyes got ya | 16:30 |
| mriedem | right, my point is it's probably there fore super old legacy reason | 16:30 |
| mriedem | like https://github.com/openstack/nova/blob/master/nova/api/openstack/compute/volumes.py#L226 | 16:30 |
| mriedem | when cinder wasn't a thing, it was nova-volume | 16:30 |
| jgriffith | mriedem +1 you're correct | 16:30 |
| jgriffith | n-vol thing | 16:30 |
| ildikov | mriedem: ok, then we need to fix that too | 16:30 |
| stvnoyes | so this show issue may have been introduced with that translate_attachments change... | 16:31 |
| stvnoyes | it was working ok earlier. | 16:31 |
| jgriffith | stvnoyes yeah, I made some changes to cinderclient that might be culprit too though. But yes it worked beautifully for a while :) | 16:32 |
| jgriffith | anyway... i think this is a Cinder show thing that needs to be fixed and I don't want to pile on things regarding changes to legacy stuff until we finish what's in flight | 16:33 |
| jgriffith | there are too many things up in the air right now IMO | 16:33 |
| ildikov | jgriffith: we still need the correct data in the volume object too | 16:33 |
| ildikov | jgriffith: so we can double check the client, we have a bit more time till the lib freeze is finally over | 16:34 |
| ildikov | anything else to the topic? | 16:35 |
| mriedem | nope | 16:36 |
| ildikov | ok, then I think that was it for today | 16:37 |
| ildikov | unless someone has further questions/comments to discuss | 16:37 |
| ildikov | ok, we can figure out topics for the joint session on the PTG offline and keep fixing up the cinderclient | 16:38 |
| ildikov | thanks everyone! | 16:38 |
| ildikov | C U here next week | 16:38 |
| hemna | ildikov, that volume_id thing I believe was a placeholder that has been there from the n-vol days, I just left it there to not break anything. | 16:38 |
| ildikov | hemna: sounds good, we have the attachment_id explicitly there, so we can fall back to that and do cleanups later | 16:39 |
| ildikov | would be my thinking re not breaking anything | 16:40 |
| ildikov | alright, have a good rest of the day everyone :) | 16:41 |
| ildikov | #endmeeting | 16:42 |
| *** openstack changes topic to "OpenStack Meetings || https://wiki.openstack.org/wiki/Meetings" | 16:42 | |
| openstack | Meeting ended Thu Aug 24 16:42:09 2017 UTC. Information about MeetBot at http://wiki.debian.org/MeetBot . (v 0.1.4) | 16:42 |
| openstack | Minutes: http://eavesdrop.openstack.org/meetings/cinder_nova_api_changes/2017/cinder_nova_api_changes.2017-08-24-16.00.html | 16:42 |
| openstack | Minutes (text): http://eavesdrop.openstack.org/meetings/cinder_nova_api_changes/2017/cinder_nova_api_changes.2017-08-24-16.00.txt | 16:42 |
| openstack | Log: http://eavesdrop.openstack.org/meetings/cinder_nova_api_changes/2017/cinder_nova_api_changes.2017-08-24-16.00.log.html | 16:42 |
| *** nhelgeson has joined #openstack-meeting-cp | 16:44 | |
| *** yamahata has joined #openstack-meeting-cp | 16:46 | |
| *** iyamahat has joined #openstack-meeting-cp | 16:47 | |
| *** mriedem has left #openstack-meeting-cp | 16:53 | |
| *** rarcea has quit IRC | 16:59 | |
| *** MarkBaker has quit IRC | 17:07 | |
| *** MarkBaker has joined #openstack-meeting-cp | 17:07 | |
| *** MarkBaker has quit IRC | 17:20 | |
| *** hemna_ has quit IRC | 18:28 | |
| *** gouthamr has quit IRC | 20:01 | |
| *** gouthamr has joined #openstack-meeting-cp | 20:04 | |
| *** gouthamr has quit IRC | 20:22 | |
| *** iyamahat has quit IRC | 20:23 | |
| *** iyamahat has joined #openstack-meeting-cp | 20:28 | |
| *** edmondsw has quit IRC | 20:43 | |
| *** gouthamr has joined #openstack-meeting-cp | 21:07 | |
| *** MarkBaker has joined #openstack-meeting-cp | 21:48 | |
| *** edmondsw has joined #openstack-meeting-cp | 21:49 | |
| *** edmondsw has quit IRC | 21:54 | |
| *** MarkBaker has quit IRC | 22:04 | |
| *** edmondsw has joined #openstack-meeting-cp | 22:52 | |
| *** nhelgeson has quit IRC | 23:18 | |
| *** _pewp_ has quit IRC | 23:48 | |
| *** _pewp_ has joined #openstack-meeting-cp | 23:52 | |
Generated by irclog2html.py 2.15.3 by Marius Gedminas - find it at mg.pov.lt!